The Kepler end-to-end data pipeline: From photons to far away worlds

Abstract

Launched by NASA on 6 March 2009, the Kepler Mission has been observing more than 100,000 targets in a single patch of sky between the constellations Cygnus and Lyra almost continuously for the last two years looking for planetary systems using the transit method. As of October 2011, the Kepler spacecraft has collected and returned to Earth just over 290 GB of data, identifying 1235 planet candidates with 25 of these candidates confirmed as planets via ground observation. Extracting the telltale signature of a planetary system from stellar photometry where valid signal transients can be as small as a 40 ppm is a difficult and exacting task. The end-to-end process of determining planetary candidates from noisy, raw photometric measurements is discussed. The Kepler mission is described in overview and the Kepler technique for discovering exoplanets is discussed. The design and implementation of the Kepler spacecraft, tracing the data path from photons entering the telescope aperture through raw observation data transmitted to the ground operations team is described. The technical challenges of operating a large aperture photometer with an unprecedented 95 million pixel detector are addressed as well as the onboard technique for processing and reducing the large volume of data produced by the Kepler photometer. The technique and challenge of day-to-day mission operations that result in a very high percentage of time on target is discussed. This includes the day to day process for monitoring and managing the health of the spacecraft, the annual process for maintaining sun on the solar arrays while still keeping the telescope pointed at the fixed science target, the process for safely but rapidly returning to science operations after a spacecraft initiated safing event and the long term anomaly resolution process. The ground data processing pipeline, from the point that science data is received on the ground to the presentation of preliminary planetary candidates and supporting data to the science team for further evaluation is discussed. Ground management, control, exchange and storage of Kepler's large and growing data set is discussed as well as the process and techniques for removing noise sources and applying calibrations to intermediate data products.